Don't worry about speakers - the F5 will take care of all of them, regardless wether they're 3-way, 2-way, full range singles or horns. The F5 is not sensitive in this regard.
Maybe you're mixing it up with the F3 that benefits from proper speakers (ideally high efficiency, high impedance).
Have fun, Hannes
Maybe you're mixing it up with the F3 that benefits from proper speakers (ideally high efficiency, high impedance).
Have fun, Hannes
test the filter on the good channel.
If you like what you hear, then keep it.
Now add the same filter to the bad amp.
Does it hide the problem?
It may be that some very high frequency grunge is coming from the source and upsetting the bad amp.
The filter is not a cure but it helps reduce sensitivity to HF if the amp needs that help.
If you like what you hear, then keep it.
Now add the same filter to the bad amp.
Does it hide the problem?
It may be that some very high frequency grunge is coming from the source and upsetting the bad amp.
The filter is not a cure but it helps reduce sensitivity to HF if the amp needs that help.
Ok, but I've tried this now with a CDP (cheapo, old) the buffalo DAC output, a Peter Daniels NOS DAC and then a phonostage all with the same effect.
I'm going to swap out the Q3 and Q4 today, I will probably remove, clean the solder pads of Q1 adn 2 as well when I'm at it adn try again with the fairchilds.
i will try your filter as well. i have an IEC inlet filter here for 6A too that I could just hook up (commecial schaffner type).
Fran
I'm going to swap out the Q3 and Q4 today, I will probably remove, clean the solder pads of Q1 adn 2 as well when I'm at it adn try again with the fairchilds.
i will try your filter as well. i have an IEC inlet filter here for 6A too that I could just hook up (commecial schaffner type).
Fran
Yeah, I know.
i'm ordering new stuff from mouser this weekend. If the last few bits I do now today don't work, I'm going to do it all again. i have a spare pair of boards.
its killing me cos the other channel is perfect.
Heres the other thing. When I had it powered up, it sounded great. So at least that keeps me going and I know its worth perservering with.
Fran
i'm ordering new stuff from mouser this weekend. If the last few bits I do now today don't work, I'm going to do it all again. i have a spare pair of boards.
its killing me cos the other channel is perfect.
Heres the other thing. When I had it powered up, it sounded great. So at least that keeps me going and I know its worth perservering with.
Fran
Fran -- when you're testing the amplifier the leads from your power supply can pick up ambient RF -- way back when this thread first started I pointed out that the amplifier could go wobbly around 600kHz -- at least it did so when I was using Pomona test leads and driving it from a PS5010 programmable supply. The ambient energy would be picked up by the test leads and excite the amplifier into oscillation.
The solution is to put a filter cap on each V+ and V- node. You can use 220uF with 100nF in parallel.
You can make a current limiter for each leg with an LM317 and LM337 -- the way that I test each and every amplifier I use is to limit the current to 400mA, adjust P1 andP2 for about 370mA, then switch to the high current supply, adjust to 1.3 A idling current and minimum DC offset without the current limiters. The PS5010 is limited to 400mA so I don't need the LM317/337's.
You should also be using 12 turn trimpots -- the adjustment can get rather twitchy at the end. With a 12 turn trimpot I can use a GR rated JFET on one leg and a BL on another and get the amplifier stable.
The solution is to put a filter cap on each V+ and V- node. You can use 220uF with 100nF in parallel.
You can make a current limiter for each leg with an LM317 and LM337 -- the way that I test each and every amplifier I use is to limit the current to 400mA, adjust P1 andP2 for about 370mA, then switch to the high current supply, adjust to 1.3 A idling current and minimum DC offset without the current limiters. The PS5010 is limited to 400mA so I don't need the LM317/337's.
You should also be using 12 turn trimpots -- the adjustment can get rather twitchy at the end. With a 12 turn trimpot I can use a GR rated JFET on one leg and a BL on another and get the amplifier stable.
Woo Hoo!!! 
OK I have it sorted!
thing is I don't really know exactly what was at fault, but prob a combination of things.
What I did was take out the board, remove the pots and Q1/2 and i cleaned all the board down removing any solder etc from where Q1/2 etc had been. I washed all the flux off the board and to be honest I couldnt see anything wrong. So I refitted the parts and reflowed all my solder joints.
I also cleaned down the heatsink and Q3/4 with alcohol and scavenged some thermal pads from a dead switch mode PS.
Fired it back up and it works. Seems very stable, bias does climb slowly and rises with temperature, but it has been playing now for about 4 hours and all is well. I let it run for 2 hours into a 6R8 power resistor to test out first.
I'm not 100% happy at leaving out the thermal compensation, as the bias does rise and fall a bit with internal temp. Perhaps someone who knows could comment on what kind of variation in bias it is normal (or at least ok) to see in use in real life, ie music playing into a speaker.
I have another question too:
I would like to get some thermal pads from mouser. Can anyone give me part number for them... I've searched and not found (for Q3/4).
Temp readings to follow later.... but safe to say all seems to be ok.
Fran
OK I have it sorted!
thing is I don't really know exactly what was at fault, but prob a combination of things.What I did was take out the board, remove the pots and Q1/2 and i cleaned all the board down removing any solder etc from where Q1/2 etc had been. I washed all the flux off the board and to be honest I couldnt see anything wrong. So I refitted the parts and reflowed all my solder joints.
I also cleaned down the heatsink and Q3/4 with alcohol and scavenged some thermal pads from a dead switch mode PS.
Fired it back up and it works.
Seems very stable, bias does climb slowly and rises with temperature, but it has been playing now for about 4 hours and all is well. I let it run for 2 hours into a 6R8 power resistor to test out first. I'm not 100% happy at leaving out the thermal compensation, as the bias does rise and fall a bit with internal temp. Perhaps someone who knows could comment on what kind of variation in bias it is normal (or at least ok) to see in use in real life, ie music playing into a speaker.
I have another question too:
I would like to get some thermal pads from mouser. Can anyone give me part number for them... I've searched and not found (for Q3/4).
Temp readings to follow later.... but safe to say all seems to be ok.
Fran
